We texted 100,000 randomly selected residents asking a simple question: How should we protect information integrity together? Thousands volunteered; 450 (random, stratified) joined an AI‑facilitated citizens’ assembly (45 rooms of ~10). Proposals included: